w a t e r r e s e a r c h 7 8 ( 2 0 1 5 ) 1 8 e2 7

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/watres

Trihalomethane hydrolysis in drinking water at elevated temperatures Xiao-lu Zhang a, Hong-wei Yang a, Xiao-mao Wang a,*, Tanju Karanfil b, Yuefeng F. Xie a,c a

State Key Joint Laboratory of Environmental Simulation and Pollution Control, School of Environment, Tsinghua University, Beijing 100084, China b Department of Environmental Engineering and Earth Sciences, Clemson University, 342 Computer Court, Anderson, SC 29625, USA c Environmental Engineering Programs, The Pennsylvania State University, Middletown, PA 17057, USA

article info

abstract

Article history:

Hydrolysis could contribute to the loss of trihalomethanes (THMs) in the drinking water at

Received 20 November 2014

elevated temperatures. This study was aimed at investigating THM hydrolysis pertaining to

Received in revised form

the storage of hot boiled water in enclosed containers. The water pH value was in the range of

21 March 2015

6.1e8.2 and the water temperature was varied from 65 to 95  C. The effects of halide ions,

Accepted 28 March 2015

natural organic matter, and drinking water matrix were investigated. Results showed that

Available online 8 April 2015

the hydrolysis rates declined in the order following CHBrCl2 > CHBr2Cl > CHBr3 > CHCl3. THM

Keywords:

tively low pH value. The activation energies for the alkaline hydrolysis of CHCl3, CHBrCl2,

hydrolysis was primarily through the alkaline pathway, except for CHCl3 in water at relaTrihalomethanes (THMs)

CHBr2Cl and CHBr3 were 109, 113, 115 and 116 kJ/mol, respectively. No hydrolysis in-

Hydrolysis

termediates could accumulate in the water. The natural organic matter, and probably other

Kinetics

constituents, in drinking water could substantially decrease THM hydrolysis rates by more

Boiling

than 50%. When a drinking water was at 90  C or above, the first order rate constants for THM

Disinfection by-products (DBPs)

hydrolysis were in the magnitude of 102‒101 1/h. When the boiled real tap water was stored in an enclosed container, THMs continued increasing during the first few hours and then kept decreasing later on due to the competition between hydrolysis and further formation. The removal of THMs, especially brominated THMs, by hydrolysis would greatly reduce one's exposure to disinfection by-products by consuming the boiled water stored in enclosed containers. © 2015 Published by Elsevier Ltd.

1.

Introduction

Chemical disinfection acts as the cornerstone unit operation of water treatment processes that secure drinking water safety (Crittenden et al., 2012; Ohar and Ostfeld, 2014). One * Corresponding author. Tel.: þ86 10 6278 1386; fax: þ86 10 6277 1472. E-mail address: [email protected] (X.-m. Wang). http://dx.doi.org/10.1016/j.watres.2015.03.027 0043-1354/© 2015 Published by Elsevier Ltd.

side effect of disinfection is the formation of disinfection byproducts (DBPs). DBPs are produced from the reactions of the disinfectants (e.g. free chlorine and chloramines) with natural organic matter (NOM) and/or bromide that are ubiquitous in natural waters (Xie, 2003; Krasner et al., 2006; Yang et al., 2012; Hua et al., 2014). Previous toxicological and

19

w a t e r r e s e a r c h 7 8 ( 2 0 1 5 ) 1 8 e2 7

epidemiological studies have reported the linkage of carcinogenic and developmental effects to exposure to DBPs in drinking water (Richardson et al., 2007). DBPs can generally be classified as either volatile or non-volatile, determined by their Henry's law constants. Trihalomethanes (THMs) and haloacetic acids (HAAs), which in most cases have the highest occurrence levels in the disinfected drinking waters and are currently regulated in a number of countries (Wang et al., 2014), belong to volatile and non-volatile DBPs, respectively. Boiling the tap water (i.e. drinking water) for the preparation of hot beverages becomes increasingly popular all around the world. Boiling could have a great effect on the DBP levels in the boiled water (Li and Sun, 2001; Krasner and Wright, 2005). Consumption of the boiled drinking water, not only cold water, should therefore be taken into consideration when collecting information of exposure to DBPs in drinking water (Krasner and Wright, 2005). Previous studies showed that the boiling effect was greatly dependent on the volatility and chemical stability of the DBP, the heating duration, and the free and combined chlorine residuals in the water. Generally, when the heating duration was sufficiently long (e.g. 5 min), the levels for THMs, trihaloacetic acids (THAAs) and several non-regulated DBPs (including halogenated aldehydes, ketones and acetonitriles) would be reduced, while the levels for dihaloacetic acids (DHAAs) would be increased or keep unchanged in chlorinated and chloraminated water, respectively. THMs and other volatile DBPs were removed mainly by volatilization, while THAAs and other non-volatile DBPs were degraded primarily by hydrolysis. The hydrolysis of THAAs and other trihalogenated DBPs (e.g. chloral hydrate) would lead to the formation of THMs (Zhang and Minear, 2002; Krasner and Wright, 2005; Hua and Reckhow, 2012). DHAAs are very stable, and the increase of DHAA levels during boiling the chlorinated water was due to further DHAA formation from both the reactions of DHAA precursors with free chlorine residuals and the degradation of intermediate DBPs (Reckhow et al., 1990). Free chlorine residuals could also react with the precursors for other DBPs (e.g. THMs), and the net effect of boiling is the sum of that on DBP loss (by volatilization and hydrolysis) and that on further formation (Liu and Reckhow, 2013; Zhang et al., 2013). A fact which has long been overlooked regarding the effect of boiling on DBP levels is that THMs can be partly removed by hydrolysis. THMs in cold drinking water are extremely stable with hydrolysis rate constants in the magnitude of 108e106

1/h (Table 1), which correspond to half-lives at hundreds to thousands of years (Mabey and Mill, 1978; Jeffers et al., 1989; Chen, 2011). However, THM hydrolysis rates can be substantially enhanced to the magnitude of 102e101 1/h when the water temperature approaches to the boiling point. It indicates that THMs can significantly be hydrolyzed in a time scale of hours. It is unusual to boil a drinking water for hours. However, in many places of the world, especially East Asia, the hot boiled water is stored in thermo bottles for many hours for the ease of preparation of hot beverages. In addition, electrical boilers were also installed in many public buildings in the world. In the enclosed containers like thermo bottles, on one hand the hydrolysis of trihalogenated DBPs would lead to the formation of THMs, on the other hand volatilization of THMs would be largely reduced. THM hydrolysis becomes important. This study was conducted to investigate the THM hydrolysis kinetics that pertains to the storage of boiled drinking water in enclosed containers. All four THMs, which are chloroform (CHCl3), bromodichloromethane (CHBrCl2), dibromochloromethane (CHBr2Cl) and bromoform (CHBr3), were included in the investigation. A water temperature range from 65 to 95  C was adopted, below which THM hydrolysis becomes negligible. The pH value of the water was slightly changed within the circum-neutral range (from 6.1 to 8.2), which can also be used to investigate the effect of water pH on the rate of THM hydrolysis. The sum of the four THM concentrations in a typical drinking water is usually in tens of mg/L level (Obolensky et al., 2007; Ding et al., 2013), and the THM speciation is critically dependent on the bromide ion concentration (Obolensky and Singer, 2005; Roccaro et al., 2014). A drinking water also contains NOM usually in mg/L level and mineral ions (e.g. chloride) usually in tens to hundreds of mg/L level. The effect of these water “impurities” on THM hydrolysis rates was also investigated. The purpose of this study is to draw one's attention to the significance of THM hydrolysis in assessing the health risks posed by DBPs contained in cool and boiled drinking water.

2.

Theory of THM hydrolysis

2.1.

THM hydrolysis pathways

The hydrolysis mechanism of chloroform, which was at very high concentrations (in g/L level), has been well studied (Hine

Table 1 e Hydrolysis rate constants (unit: 1/h) for the four THMs at various water temperatures, calculated by using the correlation equations given in previous studies. THM was either dissolved in 66.7% dioxaneewater solution at ~0.1 mol/L (Hine et al., 1956), or in water at ~10 mmol/L (Fells and Moelwynhughes, 1959) or at concentration close to saturation (1%e 10%) (Jeffers et al., 1989). The calculated hydrolysis rates for chloroform differed in different studies. Only alkaline hydrolysis rate could be obtained for CHBrCl2, CHBr2Cl and CHBr3 (Hine et al., 1956). The pH values correspond to that for 25  C. Heating temperature ( C) 25 65 75 85 90 95

CHBrCl2

CHCl3 pH 6.5

pH 7.5

pH 8.5

(0.74e2.3)  108 (1.6e2.0)  105 (7.9e8.8)  105 (2.9e3.4)  104 (5.2e6.8)  104 (0.91e1.3)  103

(0.74e1.2)  107 1.7  104 (7.2e8.0)  104 (2.5e3.4)  103 (4.4e6.6)  103 (0.74e1.3)  102

(0.74e1.1)  106 (8.8e9.0)  104 (3.2e3.6)  103 (0.99e1.4)  102 (1.6e2.5)  102 (2.6e4.6)  102

CHBr2Cl

CHBr3

pH 7.5 2.0  3.5  1.6  6.4  0.12 0.22

106 103 102 102

9.2  2.0  9.2  3.9  7.5  0.14

107 103 103 102 102

3.6  6.5  3.0  1.2  2.3  4.2 

107 104 103 102 102 102

20

w a t e r r e s e a r c h 7 8 ( 2 0 1 5 ) 1 8 e2 7

et al., 1956; Fells and Moelwyn-hughes, 1959). It was not the intention of this study to investigate the hydrolysis mechanism of THMs which are at fairly low concentrations (in tens of mg/L order). Rather, we assumed that the same hydrolysis mechanism stands, regardless of the THM species and their concentrations in water. THMs hydrolyze in two separate pathways. The neutral hydrolysis pathway (by reacting with water molecule) involves the formation of dihalomethanol, which then rapidly breaks down into carbon monoxide and formic acid (Reactions (1) and (2)). k1

CHX3 þ H2 Oƒƒƒƒ!CHX2 OH þ HX

(1)

CHX2 OH/HCOX þ HX

(2a)

ðx þ yÞHCOX þ yH2 O/xCO þ yHCOOH þ ðx þ yÞHX

(2b)

The rate-limiting step is the substitution of the first halide atom by hydroxyl group (Reaction (1)). As such, the hydrolysis rate by the neutral pathway can be described by, ðdC=dtÞn ¼ kn C ¼ k1 C;

(3)

where C is the THM concentration. The alkaline hydrolysis pathway (by reacting with hydroxide ion) involves the formation of carbon trihalide anion and carbon dihalide, the latter of which quickly breaks down into carbon monoxide and formic acid (Reactions (4e6)). k2

CHX3 þ OH ƒƒƒ ƒƒƒ ƒ! ƒ CX 3 þ H2 O k3

k4

 CX 3 ƒƒƒƒ!CX2 þ X

   Kw ¼ Hþ OH

(9)

   Ka1 ¼ Hþ H2 PO4  ½H3 PO4 

(10)

    Ka2 ¼ Hþ HPO4 2 H2 PO4 

(11)

    Ka3 ¼ Hþ PO4 3 HPO4 2

(12)

         þ  þ  þ Na þ K þ H ¼ 3 PO4 3 þ 2 HPO4 2 þ H2 PO4  þ OH (13) 

The pKw, pKa1, pKa2 and pKa3 values at 25 C are 14.00, 2.12, 7.21 and 12.67, respectively. All the above dissociation constants are dependent on the water temperature. According to the literature (Haynes, 2013), the pKw values are 12.90, 12.69, 12.50, 12.42 and 12.34 at 65, 75, 85, 90 and 95  C, respectively. The pKa values for phosphate at other temperatures were calculated by using the van't Hoff equation (Stumm and Morgan, 1996), ln

     DHf 1 K2 1 ¼  T2 T1 K1 R

(14)



(5)

whereDHf is the standard enthalpy change for the dissociation reaction, R the gas constant, T the absolute temperature and  subscripts 1 and 2 stand for the two temperatures. The DHf  2  values at 25 C for the dissociation of H3PO4, H2PO4 and HPO4 are 8.0, 4.2 and 14.7 kJ/mol, respectively. Table 2 gives the calculation results for the hydroxide concentrations. It shows that hydroxide ion concentration substantially increased with the increase of the temperature of the same phosphate buffered water.

(6)

3.

Materials and methods

3.1.

THM solutions and heating treatment

(4)

ðx þ yÞCX2 þ ð2x þ 3yÞOH /xCO þ yHCOO þ 2ðx þ yÞX þ ðx þ yÞH2 O

The equilibrium between THM and carbon trihalide anion (Reaction (4)) can be rapidly established, and the breakdown of carbon trihalide anion to carbon dihalide (Reaction (5)) is the rate-limiting step. As such, the hydrolysis rate by the alkaline pathway can be described by      ðdC=dtÞb ¼ kb OH C ¼ k2 k4 ðk3 þ k4 Þ OH C:

(7)

The observed hydrolysis rate is the sum of that for neutral hydrolysis and that for alkaline hydrolysis, i.e.   dC=dt ¼  kn þ kb OH C ¼ kobs C:

(8)

Equation (8) shows that, at a fixed water temperature, the THM hydrolysis follows pseudo-first order kinetics as long as the water pH is buffered. (Note that both kn and kb depend on the water temperature.)

2.2. Theoretical calculation for hydroxide ion concentration The pH value and the hydroxide ion concentration ([OH]) were calculated based on the water dissociation constant (Kw) and the phosphate dissociation constants (Ka1, Ka2 and Ka3) according to

The THM solutions were individually prepared by spiking an appropriate volume of each commercial concentrated solution (SigmaeAldrich, USA) into either deionized (DI) water, an aqueous solution containing sodium halides (100 mg/L chloride and 100 mg/L bromide), an aqueous solution containing humic acid (2.0 mg/L as TOC), or a drinking water matrix. In most cases, the final THM concentration was 100 mg/L; while the THM concentration was increased to 5 mmol/L when the solution was used for mass balance test. The pH of each THM solution was buffered by adding 0.4 mmol/L phosphate. In most cases, the pH was buffered at 7.5 (calculated for room temperature (25  C) and [H2PO 4 ]/ [HPO2 4 ] ¼ 33.8/66.2); while the pH value might be varied 2 ¼ 92.8/7.2), 6.9 ([H2PO among 6.1 ([H2PO 4 ]/[HPO4 ] 4 ]/  2 ] ¼ 67.0/33.0), 7.4 ([H PO ]/[HPO [HPO2 4 2 4 4 ] ¼ 39.1/60.9), 7.5, 7.8 2  2 ([H2PO 4 ]/[HPO4 ] ¼ 20.4/79.6), 8.0 ([H2PO4 ]/[HPO4 ] ¼ 13.9/  2 86.1), and 8.2 ([H2PO4 ]/[HPO4 ] ¼ 9.2/90.8) for the investigation of pH effect on hydrolysis rates. The above theoretically calculated values were basically identical to that measured by pH meter, which were 6.12, 6.89, 7.38, 7.52, 7.77, 7.98, and 8.23, respectively. The concentrations of chloride and

21

w a t e r r e s e a r c h 7 8 ( 2 0 1 5 ) 1 8 e2 7

Table 2 e Theoretically calculated hydroxide concentration (unit: 10¡6 mol/L) in the phosphate-buffered water at elevated temperatures. pH at 25  C 

T ( C)

65 75 85 90 95

6.1

6.5

6.9

7.4

7.5

7.8

8.0

8.2

8.5

0.130 0.202 0.299 0.352 0.415

0.324 0.502 0.744 0.876 1.03

0.813 1.26 1.86 2.18 2.56

2.54 3.88 5.67 6.62 7.72

3.14 4.79 6.96 8.10 9.43

5.98 8.90 12.6 14.4 16.5

8.72 12.6 17.3 19.6 22.2

11.9 16.4 21.7 24.3 27.2

17.7 23.3 29.5 32.4 35.6

bromide ions, and humic acid, if added, were typical of that in drinking water, and were added to investigate the effects of halide ions and natural organic matter, respectively. The drinking water matrix was a grab sample collected from a local water treatment plant at the point just prior to primary disinfection. The plant treated raw surface water following a treatment process that consisted of prechlorination, clarification, sand filtration, granular activated carbon filtration and primary disinfection. Therefore, the collected water sample was not a finished drinking water but a wellpretreated raw surface water. It had a TOC concentration at 2.8 mg/L, UV254 absorbance at 0.016 1/cm, and nondetectable residual chlorine. Very low DBP levels were detected in the water sample before THM spiking. Each THM solution was dispensed into a number of amber glass vials (40 mL in volume, each equipped with a Teflon septa and a screw cap) with no headspace left. In this way, the THM volatilization from the solution could be effectively prevented. The vials were then transferred into a hot water bath. The water bath temperature was varied among 65, 75, 85, 90 and 95  C for the investigation of the temperature effect on hydrolysis kinetics. The heating duration was up to 216 h, depending on the hydrolysis rate. After a preset heating duration, the vials were quickly transferred into a cool water bath at 20  C. The cooled samples were then determined for the residual THM concentration and, if required, the formed halide concentrations. All heating treatment and THM analysis were conducted in duplicate for each solution and condition.

3.2.

Boiling and storage of real tap water

A real tap water sample was collected from a household tap located at Clemson, SC, USA. The collected water sample had a pH at 7.1, TOC concentration at 1.0 mg/L, UV254 absorbance at 0.016 1/cm, and residual chlorine at 1.0 mg/L. The concentrations of DBPs (including THMs, HAAs and other common species) in the water sample were measured as the already formed DBPs levels before heating treatment. The tap water was fast heated (in 6 min) to just boiling in an electrical kettle. A volume of the boiled water was quickly cooled down to 20  C for the determination of the DBP levels after boiling. The remaining boiled water was quickly dispensed into a number of 40 mL amber glass vials with no headspace left, which were then immediately transferred to a water bath set at 90  C. They were kept at that temperature for up to 72 h. After preset storage durations, appropriate vials were quickly cooled down to 20  C. The DBP levels were then determined to investigate the effect of storage.

3.3.

Analytical methods

Both THMs and HAAs were analyzed by using a gas chromatograph equipped with an electron capture detector (GC 7890A, Agilent, USA) after the micro liquideliquid extraction by using methyl tert-butyl ether as the extraction agent (and, additionally, esterification for HAAs by using diazomethane). More details could be found by referring to the USEPA Method 551.1 (1995) and APHA et al (2005), respectively. The detection limits were approximately 0.1 mg/L for each THM, chloral hydrate (CH), dichloroacetonitrile (DCAN) and 1,1,1-trichloropropanone (TCPN), and 0.5 mg/L for each HAA, respectively. Chloride and bromide were analyzed by using an ion chromatograph (761, Metrohm, Switzerland). The detection limit was 5 mg/L for both chloride and bromide. The TOC and UV254 were analyzed by using a TOC analyzer (TOC-V CPH, Shimadzu, Japan) and an ultraviolet spectrophotometer (UV1800, Shanghai Aoxi, China), respectively.

4.

Results and discussions

4.1.

THM hydrolysis at elevated water temperatures

THM was dissolved in DI water and pH-buffered at 6.1, 6.9, 7.5, 7.8, 8.0 or 8.2 by using 0.4 mmol/L phosphate. The THM solution was then heated at 65, 75, 85, 90 or 95  C. The hydrolysis kinetics was analyzed by plotting the THM concentration (C) versus heating time (t) and fitted by using Equation (8) (Fig. 1). Because hydroxide ion concentration would not change during heating, it was expected that the THM hydrolysis followed the pseudo-first order kinetics, as was the case for all four THMs under all conditions investigated. The obtained hydrolysis rate constants (kobs) (Table 3) were compared with that calculated by using correlation equations given in previous studies (Table 1) (Hine et al., 1956; Fells and Moelwynhughes, 1959; Jeffers et al., 1989). It appears that initial THM concentration had minimal effect on the hydrolysis rate. The THM concentration was 100 mg/L in this study, while it was from g/L to tens of g/L in previous studies. The hydrolysis rate constant determined when the THM concentration was low (in this study) was only slightly higher than that determined when the THM concentration was high (in previous studies). The pH value of the THM solution (calculated for and measured at 25  C) was varied to obtain the neutral (kn) and the alkaline (kb) hydrolysis rate constants for the four THMs at both 90  C and 65  C (Equation (8)). As expected, the observed hydrolysis rate constant (kobs) for each THM was in a linear

22

w a t e r r e s e a r c h 7 8 ( 2 0 1 5 ) 1 8 e2 7

2.71  104, 1.81  104 and 7.36  103 L/mol h, respectively; while the neutral hydrolysis rate constants (kn) were all about 1  103 1/h. At 65  C, the kb values for CHCl3, CHBrCl2, CHBr2Cl and CHBr3 were 1.81  102, 1.17  103, 8.45  103 and 3.58  102 L/mol h, respectively; while the kn values could not be accurately determined but were generally in the magnitude of 105‒104 1/h. It indicated that, except for CHCl3, under typical pH ranges of drinking water (e.g. 6.5e8.5), alkaline hydrolysis would predominate over neutral hydrolysis when the water temperature was within the range from 65 to 90  C. For example, if the CHBrCl2 solution of a pH value at 6.5 was heated to 65 and 90  C, the ratios of the (pseudo-) first-order rate constant for alkaline hydrolysis to that for neutral hydrolysis of CHBrCl2 were about 30 and 10, respectively, while they were about 2000 and 500, respectively, if the pH value of the CHBrCl2 solution was 8.5. However, for CHCl3, alkaline hydrolysis could predominate over neutral hydrolysis only when the solution pH value was sufficiently high (e.g. the ratios were about 10 and 3, respectively, when the pH was 7). However, it was assumed that THM hydrolysis was all caused by the alkaline pathway hereafter. The kb values obtained under the various temperatures (from 65 to 95  C) were correlated by using the Arrhenius equation, k ¼ A expð  Ea =RTÞ

Fig. 1 e THM concentration variation with time in the water that was at (a) 75  C and (b) 95  C. The buffered DI water had a pH of 7.5 at 25  C.

relationship with the hydroxide ion concentration (Fig. 2). The hydroxide ion concentrations could be found in Table 2. At 90  C, the alkaline hydrolysis rate constants (kb) for CHCl3, CHBrCl2, CHBr2Cl and CHBr3 were determined to be 1.65  103,

(15)

The activation energies (Ea) were determined to be 109, 113, 115 and 116 kJ/mol for CHCl3, CHBrCl2, CHBr2Cl and CHBr3, respectively (Fig. 3). The activation energies were unusually high, indicating the strong dependence of THM hydrolysis rate on water temperature. In addition, the activation energies only differed slightly among the four THMs. According to the alkaline hydrolysis mechanisms, the breakdown of CX 3 to CX2 (Reactions (5)) is the rate-limiting step. It could be assumed that the difference of the activation energies among the four THMs might be due to the different strength of CeX bond (C stands for carbon, X stands for halogen). The bond strengths follow the order CHCl3> CHBr3> CHBrCl2z CHBr2Cl (note that the CeBr bond is weaker than the CeCl bond for the

Table 3 e The determined hydrolysis rate constants (unit: 1/h) for the four THMs of trace amount in the water at elevated temperatures and typical drinking water pH values. Water matrix (all buffered by phosphate) DI water

100 mg/L chloride and 100 mg/L bromide solution 2.0 mg/L (as TOC) humic acid solution Drinking water matrix

Heating T ( C)

pH (at 25  C)

65 65 65 65 75 85 90 90 90 90 90 95 90

7.5 7.8 8.0 8.2 7.5 7.5 6.1 6.9 7.5 7.8 8.0 7.5 7.5

1.2 1.7 2.0 1.6 6.7 3.4 4.9 1.4 2.7 3.4 2.3 1.3

90 90

7.5 7.5

7.8  103 3.5  103

CHCl3 e  103  103  103  102  103  103  103  102  102  102  102  102

CHBrCl2 4.1  7.0  1.0  1.4  1.7  8.0  1.1  5.5  0.15 0.34 0.55 0.33 0.15

103 103 102 102 102 102 102 102

6.1  102 3.1  102

CHBr2Cl 2.7  5.0  7.6  9.9  1.2  5.8  8.0  4.1  0.11 0.23 0.37 0.23 0.12

103 103 103 103 102 102 103 102

4.5  102 2.3  102

CHBr3 1.1  1.9  3.1  4.4  5.2  2.7  3.8  1.8  4.5  9.3  0.15 9.3  4.7 

103 103 103 103 103 102 103 102 102 102 102 102

2.0  102 1.1  102

w a t e r r e s e a r c h 7 8 ( 2 0 1 5 ) 1 8 e2 7

23

hydrolysis mechanisms (Reactions (4e6)), the concentration of carbon trihalide anion (CX 3 ) could be a critical factor in controlling the hydrolysis rate. A higher carbon trihalide anion concentration corresponded to a higher hydrolysis rate (and a larger pre-exponential factor value). After the breakdown of CX 3 to CX2, the ensuing steps in which CX2 hydrolyzes to carbon monoxide and formic acid are very rapid (Reaction (6)). Complete dehalogenation could therefore be expected along with the loss of THMs. Mass balance for halide ions and THMs were established as C0  C ¼

      Cl þ Br 3

(16)

where all species are in molar concentration, C0 is the initial THM concentration before heating, and C is the measured THM concentration after heating. The initial THM concentration was set at 5 mmol/L for the better measurement of the released halogen ions. Results showed that Equation (16) stood for all the four THMs (Fig. 4). The verification of this hydrolysis pathway is important under the drinking water condition. It indicates that no potentially hazardous halogencontaining intermediates would accumulate during the hydrolysis of any THM.

4.2.

Fig. 2 e Linear correlation of the pseudo-first order rate constant for THM hydrolysis to the hydroxide ion concentration. The THM-containing water was at (a) 90  C and (b) 65  C, respectively.

brominated THMs) but in actuality differed only slightly (Haynes, 2013). In contrast, the pre-exponential factors (A) differed greatly among CHCl3, CHBrCl2, CHBr2Cl and CHBr3, which were 8.03  1018, 3.46  1020, 5.61  1020 and 2.54  1020 L/mol h, respectively. According to the alkaline

Fig. 3 e Correlation of the rate constants for THM alkaline hydrolysis (kb) with the water temperatures.

Effects of water chemistry on THM hydrolysis

Typical drinking water is a solution that contains natural organic matter (NOM) in mg/L level, chloride ion in tens to hundreds of mg/L level, bromide ion in tens to hundreds of mg/ L level, and other molecules and ions in various levels. In contrast, the THM concentrations are in tens of mg/L level. Therefore, it is important to investigate the effect of water matrix on the THM hydrolysis. Reaction (5) could be reversible, and the increase of halide ions might slow down the THM hydrolysis rate. The results by dosing 100 mg/L chloride and 100 mg/L bromide into the phosphate-buffered deionized water indicated that halide concentrations at the typical levels in drinking water had negligible effect on THM hydrolysis rates (Fig. 5). The differences were within 5% for all four THMs when the THM solution was heated to 90  C. A plausible reason was that the rate constant for the reversal of Reaction (5) was very low compared with that for the forward reaction. To explore the effects of NOM on THM hydrolysis rates, commercial humic acids were dosed into DI water to a final TOC concentration at 2.0 mg/L, which was then pH-buffered at 7.5 (25  C). Humic substances are the major organic components in most natural surface waters. Results demonstrated that humic acid could substantially decrease the hydrolysis rates for all THMs (Fig. 5). When the water contained 2.0 mg/L humic acids as TOC (pH ¼ 7.5, 25  C), the hydrolysis rate constants for CHCl3, CHBrCl2, CHBr2Cl and CHBr3 at 90  C were decreased to 0.0078, 0.0612, 0.0453 and 0.0198 1/h, respectively, by 55% on average (Table 3). According to the hydrolysis mechanisms and Equation (8), the plausible reason for the reduction of hydrolysis rate was the decrease of the hydroxide ion concentration in solution when humic acids were present under otherwise identical conditions. Humic acids are weak acids which could also participate in the acid‒base equilibria as Equations 9e13 show.

24

w a t e r r e s e a r c h 7 8 ( 2 0 1 5 ) 1 8 e2 7

Fig. 4 e Mass balance for (a) CHCl3, (b) CHBrCl2, (c) CHBr2Cl and (d) CHBr3 hydrolysis in the water at 90  C.

Fig. 5 e Hydrolysis of (a) CHCl3, (b) CHBrCl2, (c) CHBr2Cl and (d) CHBr3 in the various water matrices (all buffered by phosphate) that was at 90  C.

In addition to halide ions and humic substances, drinking water also contains other mineral ions, dissolved organic substances and colloidal particles. The THM hydrolysis in a locally sampled drinking water matrix was consequently investigated. The collected water sample was a wellpretreated raw surface water but with no primary disinfection (and as such contained non-detectable residual chlorine and negligible background DBPs) and had a pH value of 7.6, which was phosphate-buffered to 7.5 (at 25  C) and spiked with each THM at 100 mg/L before heated at 90  C. Similar to the effect of humic acids, the hydrolysis rate constants for CHCl3, CHBrCl2, CHBr2Cl and CHBr3 at 90  C were decreased to

0.0035, 0.0309, 0.0228, and 0.0111 1/h, respectively, by 77% on average (Fig. 5) (Table 3). A direct comparison of the hydrolysis rate constants in the humic acid-containing water and the real drinking water might not be appropriate. Though the effect of NOM on THM hydrolysis kinetics was evident, the effects of other components in the drinking water could not be excluded. In addition to NOM, drinking water also contains a number of dissolved chemicals, colloids, microorganism metabolites, suspended solids and etc. Moreover, the NOM properties of various water sources could be very different (Reckhow et al., 1990). The specific effects of various drinking

25

w a t e r r e s e a r c h 7 8 ( 2 0 1 5 ) 1 8 e2 7

water ingredients on THM hydrolysis rates need to be further investigated.

4.3. water

The significance of THM hydrolysis for drinking

Tap water normally contains the already formed DBPs, DBP precursors and disinfectant residual. During heating of tap water to the boiling point, the THM concentrations might substantially increase due to the further and enhanced reactions between the precursors and free chlorine and the enhanced hydrolysis of trihalogenated DBPs, such as trihaloacetic acids, trihaloacetaldehydes and trihalopropanones (Zhang and Minear, 2002; Hua and Reckhow, 2012; Pan and Zhang, 2013; Zhang et al., 2013; Liu and Reckhow, 2013). The extent of concentration increase depends on the free chlorine level in the heated tap water (Li and Sun, 2001; Krasner and Wright, 2005). For the particular example of this study, during heating of the real tap water (with a free chlorine residual at 1.0 mg/L) to the boiling point, the CHCl3, CHBrCl2 and CHBr2Cl concentrations were increased from 7.3 to 14.0 mg/L, from 4.2 to 6.1 and from 2.0 to 2.7 mg/L, respectively (Table 4). (No CHBr3 was detected in either water sample.) Simultaneously, the TCPN concentration was decreased from 0.9 to 0.4 mg/L, while the TCAA and CH concentrations were increased from 4.9 to 5.9 mg/L and from 3.6 to 5.3 mg/L, respectively. (All DHAA and DCAN concentrations were increased.) Although THMs are volatile compounds (the Henry's constants at 95  C for CHCl3, CHBrCl2, CHBr2Cl and CHBr3 are 57.9, 28.1, 20.5 and 10.9 L∙bar/mol, respectively), volatilization of THMs into the atmosphere is only favored by continuous agitation (e.g. by boiling). Therefore, in practice, the THM concentration remains quite high in the boiled water, unless the water is let for continuous boiling for at least several minutes. If the hot boiled water were then stored in an enclosed container for sufficient long time, the THM concentrations would be firstly increased and then decreased. In such an enclosed container, THM volatilization into the atmosphere is greatly reduced. Nevertheless, on one hand, THM can be further formed, e.g. from hydrolysis of the above-mentioned trihalogenated DBPs. The pseudo-first order rate constants (at 95  C) for the hydrolysis of bromodichloroacetic acid,

dibromochloroacetic acid and tribromoacetic acid were determined to be ca. 5.2, 24.9 and 27.6 1/h, respectively (Zhang and Minear, 2002). On the other hand, THM hydrolysis could be substantial. The hydrolysis rate constants determined in this study for the four THMs in a typical drinking water at or near the boiling point were in the magnitude of 102‒101 1/h (Table 3). For the particular example of this study, keeping the boiled tap water at 90  C led to the depletion of TCPN, CH and TCAA in 0.5, 2 and 8 h, respectively. In the mean time, the concentrations for CHCl3, CHBrCl2 and CHBr2Cl continued increasing during the first 8 h and then kept decreasing later on (Table 4). This way of hot water storage is therefore effective to remove trihalogenated DBPs including THMs from the water. Brominated DBPs are believed more toxic than their chlorinated analogs. This study showed that under otherwise identical conditions, CHBrCl2 and CHBr2Cl hydrolyze in a faster rate than CHBr3 and CHCl3. The THM speciation in chlorinated waters depends on both the bromide concentration and the chlorine dosage. However, unless bromide concentration is very high, CHCl3 and CHBrCl2 are normally the major THM species. Therefore, when the conditions are favorable for THM hydrolysis, the THM that eventually remains in the highest level should be CHCl3, which is the least toxic among the four regulated THMs. For the particular example of this study, in the first 8 h, the concentration increase rate for CHCl3 was much higher than that for CHBrCl2 and CHBr2Cl. After 8 h of storage, the concentration decrease rate for CHCl3 was much lower than that for CHBrCl2 and CHBr2Cl. Both observations could be partly explained by the higher hydrolysis rates of CHBrCl2 and CHBr2Cl than CHCl3. (One other possible explanation was the slower formation rates of CHBrCl2 and CHBr2Cl than CHCl3 when bromide ion concentration was low.) In addition, the limiting step of both neutral and alkaline hydrolysis is the cleavage of the first halogen atom (Reactions (1) and (5)). It indicates that no potentially toxic intermediates will accumulate after THM hydrolysis. Heating water therefore has an additional impact on reducing the health effect posed by THMs, although toxicity (e.g. cytotoxicity and genotoxicity) measurement need be conducted for a better illustration. If the boiled water were let open to the atmosphere, the water temperature would gradually decrease to the room temperature in a couple of hours. In this case, THM hydrolysis

Table 4 e Variation of DBP concentrations (unit: mg/L) in real tap water caused by boiling and the ensuing storage at 90  C for different durations. DBPs

CHCl3 CHBrCl2 CHBr2Cl DCAA BCAA DBAA TCAA CH TCPN DCAN

Kept at 90  C for

Before boiling

7.3 4.2 2.0 1.9 3.0 1.8 4.9 3.6 0.9 0.4

0h

0.5 h

2h

4h

8h

16 h

24 h

48 h

72 h

14.0 6.1 2.7 3.5 4.0 2.0 5.9 5.3 0.4 0.6

29.5 9.2 3.4 3.9 4.0 1.9 4.1 1.2 0.0 0.4

29.3 8.2 3.1 3.3 3.8 2.1 2.7 0.0

36.1 9.2 3.1 3.3 3.4 1.8 1.9

42.3 9.8 3.5 3.7 3.6 1.8 0.0

40.1 7.3 3.0 3.6 3.4 1.7

34.7 7.4 2.8 3.2 3.1 1.7

26.2 3.1 1.7 3.2 2.6 1.7

/

0.0

1.7 1.3 2.6 2.2 1.6

26

w a t e r r e s e a r c h 7 8 ( 2 0 1 5 ) 1 8 e2 7

can be negligible, considering the short time period and the lowered water temperature. For example, the hydrolysis rate constants at 65  C for the four THMs in a typical drinking water are in the magnitude of 104‒103 1/h (Table 3), indicating the half-lives ranged from days to months. However, some water utilities adjust the pH of their finished water a bit higher in order to comply with the Lead and Copper Rule (USEPA, 2007). At alkaline conditions, the THM hydrolysis is substantially enhanced. In a water at pH 9.0 (measured at 25  C) and 65  C, the THM hydrolysis rate constants could be in the magnitude of 102e101 1/h and THM hydrolysis becomes considerable. A number of epidemiological studies have been conducted to investigate the exposure risks posed by DBPs. It is wellknown that the DBP concentrations vary spatially and temporally in the water distribution systems (Williams et al., 1997; Rodriguez et al., 2004; Symanski et al., 2004). Attentions were also paid to the DBP variations in the plumbing systems and the residential hot water tanks (Dion-Fortier et al., 2009; Chowdhury et al., 2011). The effect of boiling water on THM levels was also investigated, focusing on further formation (from the precursors and the decomposition of other DBPs) and volatilization of the THMs (Li and Sun, 2001; Wu et al., 2001; Krasner and Wright, 2005). Recently, the heating effect on DBP levels in the waters of different water ages was systematically investigated by taking into account the disinfectant residual variation with water age in water distribution systems (Liu and Reckhow, 2013). The current study stepped further into the THM hydrolysis in the heated water stored for a relatively long duration. The results might help raise water professionals' concerns on the fate and transformation of THMs and other conceived thermally stable DBPs in hot drinking waters when conducting epidemiological studies and regulatory estimations.

5.

Conclusions

Hydrolysis of the four regulated THMs pertaining to storage of hot boiled drinking water in enclosed containers was investigated by varying the water temperature and the water pH value and by using various water matrices. It revealed that all four THMs underwent considerable hydrolysis in the water at the temperature range from 65 to 95  C with an initial pH within 6.1 and 8.2. When the water was pH buffered, the THM hydrolysis followed pseudo-first order kinetics, regardless of the water pH, temperature and matrix. Under otherwise identical conditions, the hydrolysis rates showed a declining order following CHBrCl2 > CHBr2Cl > CHBr3 > CHCl3. At circum-neutral pH ranges for most drinking water, the alkaline pathway predominated over the neutral pathway, except for CHCl3 in water of relatively low pH value. Regarding the alkaline pathway, cleavage of the first halogen atom was the rate-limiting step. The activation energies for the alkaline hydrolysis of the four THMs were unusually high and differed only slightly, which were 109, 113, 115 and 116 kJ/mol for CHCl3, CHBrCl2, CHBr2Cl and CHBr3, respectively. The difference of THM hydrolysis rates could be due to its reactivity with hydroxide ion and the product's reactivity with water. The normal occurrence level of halogen anions (tens to hundreds

of mg/L for Cl, tens of mg/L for Br) could not affect THM hydrolysis kinetics. The commercial humic substances (TOC ¼ 2 mg/L) and natural organic matter (and probably other constituents) in drinking water substantially decreased THM hydrolysis rates, by 55% and 77% on average, respectively. It was likely due to the reduction of hydroxide ion concentration by the enhanced ionization of organic acids and by carbonate precipitation with cations at elevated temperature in drinking water matrix. When a drinking water was at 90  C or above, the THM hydrolysis rate constants were all in the magnitude of 102‒101 1/h, which indicated that all four THMs could be largely removed by hydrolysis in the magnitude of hours. When the boiled real tap water was stored in an enclosed container at 90  C, THMs continued increasing during the first 8 h and then kept decreasing later on due to the competition between hydrolysis and further formation. The least stable CHBrCl2 and CHBr2Cl could be almost eliminated after storage for 3 d. The benefit margin of boiling/storing drinking waters was unclear so far unless the overall DBP toxicity was considered. In the situation while volatilization of THMs is largely blocked, the role of THM hydrolysis can be significantly enhanced. These results on THM hydrolysis at elevated water temperatures could help water professionals to better assess the health risks posed by DBPs due to cool and boiled tap water consumption.

Acknowledgments The authors acknowledge the financial support provided by the National Natural Science Foundation of China (No. 51290284 and No. 51278268). The authors want to thank Wilson Beita-Sandi from Rich Lab in Clemson University for his assistance in HAA sample analysis.

references

APHA, AWWA, WEF, 2005. Standard Methods for the Examination of Water and Wastewater, 21st ed. Washington, DC. Chen, B., 2011. Hydrolytic stabilities of halogenated disinfection byproducts: review and rate constant quantitative structureeproperty relationship analysis. Environ. Eng. Sci. 28 (6), 385e394. Chowdhury, S., Rodriguez, M.J., Sadiq, R., Serodes, J., 2011. Modeling DBPs formation in drinking water in residential plumbing pipes and hot water tanks. Water Res. 45 (1), 337e347. Crittenden, J.C., Trussell, R.R., Hand, D.W., Howe, K.J., Tchobanoglous, G., 2012. MWH's Water Treatment: Principles and Design, third ed. John Wiley & Sons, Hoboken, NJ. Ding, H.H., Meng, L.P., Zhang, H.F., Yu, J.W., An, W., Hu, J.Y., Yang, M., 2013. Occurrence, profiling and prioritization of halogenated disinfection by-products in drinking water of China. Environ. Sci. Process. Impact 15 (7), 1424e1429. rodes, J., Proulx, F., 2009. Dion-Fortier, A., Rodriguez, M.J., Se Impact of water stagnation in residential cold and hot water plumbing on concentrations of trihalomethanes and haloacetic acids. Water Res. 43 (12), 3057e3066. Fells, I., Moelwynhughes, E.A., 1959. The kinetics of the hydrolysis of the chlorinated methanes. J. Chem. Soc. 398e409.

w a t e r r e s e a r c h 7 8 ( 2 0 1 5 ) 1 8 e2 7

Haynes, W.M., 2013. CRC Handbook of Chemistry and Physics, 94th ed. CRC Press, Boca Raton, FL. Hine, J., Dowell, A.M., Singley, J.E., 1956. Carbon dihalides as intermediates in the basic hydrolysis of haloforms. 4. Relative reactivities of haloforms. J. Am. Chem. Soc. 78, 479e482. Hua, G.H., Kim, J., Reckhow, D.A., 2014. Disinfection byproduct formation from lignin precursors. Water Res. 63, 285e295. Hua, G., Reckhow, D.A., 2012. Effect of alkaline pH on the stability of halogenated DBPs. J. Am. Water Work Assoc. 104 (2), 49e50. Jeffers, P.M., Ward, L.M., Woytowitch, L.M., Wolfe, N.L., 1989. Homogeneous hydrolysis rate constants for selected chlorinated methanes, ethanes, ethenes, and propanes. Environ. Sci. Technol. 23 (8), 965e969. Krasner, S.W., Weinberg, H.S., Richardson, S.D., Pastor, S.J., Chinn, R., Sclimenti, M.J., Onstad, G.D., Thruston Jr., A.D., 2006. Occurrence of a new generation of disinfection byproducts. Environ. Sci. Technol. 40 (23), 7175e7185. Krasner, S.W., Wright, J.M., 2005. The effect of boiling water on disinfection by-product exposure. Water Res. 39 (5), 855e864. Li, X.Z., Sun, J.M., 2001. Further formation of trihalomethanes in drinking water during heating. Int. J. Environ. Health Res. 11 (4), 343e348. Liu, B., Reckhow, D.A., 2013. DBP formation in hot and cold water across a simulated distribution system: effect of incubation time, heating time, ph, chlorine dose, and incubation temperature. Environ. Sci. Technol. 47 (20), 11584e11591. Mabey, W., Mill, T., 1978. Critical review of hydrolysis of organic compounds in water under environmental conditions. J. Phys. Chem. Ref. Data 7 (2), 383e415. Obolensky, A., Singer, P.C., 2005. Halogen substitution patterns among disinfection byproducts in the information collection rule database. Environ. Sci. Technol. 39 (8), 2719e2730. Obolensky, A., Singer, P.C., Shukairy, H.M., 2007. Information collection rule data evaluation and analysis to support impacts on disinfection by-product formation. J. Environ. Eng 133 (1), 53e63. Ohar, Z., Ostfeld, A., 2014. Optimal design and operation of booster chlorination stations layout in water distribution systems. Water Res. 58, 209e220. Pan, Y., Zhang, X., 2013. Four groups of new aromatic halogenated disinfection byproducts: effect of bromide concentration on their formation and speciation in chlorinated drinking water. Environ. Sci. Technol. 47 (3), 1265e1273. Reckhow, D.A., Singer, P.C., Malcolm, R.L., 1990. Chlorination of humic materials - by-product formation and chemical interpretations. Environ. Sci. Technol. 24 (11), 1655e1664. Richardson, S.D., Plewa, M.J., Wagner, E.D., Schoeny, R., Demarini, D.M., 2007. Occurrence, genotoxicity, and

27

carcinogenicity of regulated and emerging disinfection byproducts in drinking water: a review and roadmap for research. Mutat. Res. Rev. Mutat. Res. 636 (1e3), 178e242. Roccaro, P., Korshin, G.V., Cook, D., Chow, C.W.K., Drikas, M., 2014. Effects of pH on the speciation coefficients in models of bromide influence on the formation of trihalomethanes and haloacetic acids. Water Res. 62, 117e126. Rodriguez, M.J., Serodes, J.B., Levallois, P., 2004. Behavior of trihalomethanes and haloacetic acids in a drinking water distribution system. Water Res. 38 (20), 4367e4382. Stumm, W., Morgan, J.J., 1996. Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters. Wiley, New York, NY. Symanski, E., Savitz, D.A., Singer, P.C., 2004. Assessing spatial fluctuations, temporal variability, and measurement error in estimated levels of disinfection byproducts in tap water: implications for exposure assessment. Occup. Environ. Med. 61 (1), 65e72. USEPA, 1995. Determination of Chlorination Disinfection Byproducts, Chlorinated Solvents, and Halogenated Pesticides/ Herbicides in Drinking Water by Liquideliquid Extraction and Gas Chromatography with Electron-capture Detection (Method 551.1). Office of Research and Development, Cincinnati, Ohio. USEPA, 2007. National primary drinking water regulations for Lead and copper: short-term regulatory revisions and clarifications. Fed. Regist. 72, 57782e57820. Wang, X.M., Mao, Y.Q., Tang, S., Yang, H.W., Xie, Y.F., 2014. Disinfection byproducts in drinking water and regulatory compliance: a critical review. Front. Environ. Sci. Eng. http:// dx.doi.org/10.1007/s11783-014-0734-1. Williams, D.T., LeBel, G.L., Benoit, F.M., 1997. Disinfection byproducts in Canadian drinking water. Chemosphere 34 (2), 299e316. Wu, W., Benjamin, M.M., Korshin, G.V., 2001. Effects of thermal treatment on halogenated disinfection by-products in drinking water. Water Res. 35 (15), 3545e3550. Xie, Y.F., 2003. Disinfection Byproducts in Drinking Water: Formation, Analysis, and Control. CRC Press, Washington, DC. Yang, X., Shen, Q.Q., Guo, W.H., Peng, J.F., Liang, Y.M., 2012. Precursors and nitrogen origins of trichloronitromethane and dichloroacetonitrile during chlorination/chloramination. Chemosphere 88 (1), 25e32. Zhang, X., Minear, R.A., 2002. Decomposition of trihaloacetic acids and formation of the corresponding trihalomethanes in drinking water. Water Res. 36 (14), 3665e3673. Zhang, X.L., Yang, H.W., Wang, X.M., Fu, J., Xie, Y.F., 2013. Formation of disinfection by-products: effect of temperature and kinetic modeling. Chemosphere 90 (2), 634e639.

Trihalomethane hydrolysis in drinking water at elevated temperatures.

Hydrolysis could contribute to the loss of trihalomethanes (THMs) in the drinking water at elevated temperatures. This study was aimed at investigatin...
782KB Sizes 2 Downloads 6 Views